- input
- $0.111
- cached
- $0.097
- output
- $4.44
- Context
- 1Mtokens
- Released
- 3mo ago
- Size
- 1.7Tparameters
Cost/1M tokens
Benchmarks
DeepSWE is the primary coding benchmark where an exact model evaluation exists. Arena leaderboards and other sourced evaluations appear below; missing scores are not zero.
How it ranks
Against its predecessor, similarly priced models and the models ranked around it.
DeepSWE 1.1
Best pass@1 per model · Higher is better
Not evaluated in DeepSWE 1.1
Arena Text
Best Arena rating per model · Higher is better
Arena #59
- 1466
Grok 4.5Nearby rank - 1465
GLM 5.1Nearby rank - 1464
DeepSeek V4 Pro 0813 - 1461
Claude Sonnet 5Nearby rank - 1461
Kimi K2.6Similar price - 1426
Nemotron 3 UltraSimilar price - 1423
Mistral Large 4Similar price - 1400
Qwen3 235B A22B Thinking 2507Similar price
Ahead of 3 of 4 similarly priced models
Arena WebDev
Best Arena rating per model · Higher is better
Arena #44
- 1632
Hy4 previewSimilar price - 1583
Gemini 3.8 FlashNearby rank - 1582DeepSeek V4 Flash 0423Nearby rank
- 1581DeepSeek V4 Pro 0813
- 1577
GPT-6 LunaNearby rank - 1573
Step 5 PreviewNearby rank - 1564Mistral Large 4Similar price
- 1474Kimi K2.7 CodeSimilar price
- 1400Qwen3.5 397B A17BSimilar price
Ahead of 2 of 4 similarly priced models
Arena Search
Best Arena rating per model · Higher is better
Not ranked on Arena Search yet
Agent Arena
Best IPS score per model · Higher is better
Not ranked on Agent Arena yet
ARC-AGI-2
Best score per model · Higher is better
#33 of 85
- 65.1%Grok 4.20Similar price
- 61.4%Grok 4.7Nearby rank
- 61.3%Claude Haiku 5.5Nearby rank
- 61.3%DeepSeek V4 Pro 0813
- 60.4%Claude Sonnet 4.6Nearby rank
- 60.4%Gemini 3.6 FlashNearby rank
- 11.8%Kimi K2.5Similar price
- 4.9%GLM 5Similar price
- 1.3%Gemini 3 Flash PreviewSimilar price
Ahead of 3 of 4 similarly priced models
FrontierMath-Tier-4-v2-Private
Best score per model · Higher is better
#38 of 72
- 29.3%GLM 5.3Nearby rank
- 26.8%Claude Opus 4.6Nearby rank
- 26.8%DeepSeek V4 Pro 0813
- 26.8%DeepSeek V4.1 FlashNearby rank
- 26.8%Gemini 3.1 Pro PreviewNearby rank
- 26.8%Mistral Large 4Similar price
- 17.1%Gemini 3 Flash PreviewSimilar price
- 14.6%Grok 4.3Similar price
- 12.2%Kimi K2.7 CodeSimilar price
Ahead of 3 of 4 similarly priced models
FrontierMath-Tiers-1-3-v2-Private
Best score per model · Higher is better
#38 of 87
- 65.6%Claude Sonnet 5Nearby rank
- 65.6%Qwen3.8 Max (0902)Nearby rank
- 64.6%DeepSeek V4 Pro 0813
- 64.6%Qwen3.7 MaxNearby rank
- 62.8%Gemini 3.5 FlashNearby rank
- 58.6%Mistral Large 4Similar price
- 54.0%Kimi K2.7 CodeSimilar price
- 51.2%Gemini 3 Flash PreviewSimilar price
- 31.2%Qwen3.5 397B A17BSimilar price
Ahead of 4 of 4 similarly priced models
GPQA diamond
Best score per model · Higher is better
#28 of 161
- 92.3%Qwen3.8 Max (0902)Nearby rank
- 91.9%GLM 5.2Nearby rank
- 91.7%DeepSeek V4 Pro 0813
- 91.6%GPT-5.6 LunaNearby rank
- 91.4%GPT-5.2Nearby rank
- 89.4%Gemini 3 Flash PreviewSimilar price
- 88.1%Mistral Large 4Similar price
- 87.9%Kimi K2.7 CodeSimilar price
- 80.1%Qwen3 235B A22B Thinking 2507Similar price
Ahead of 4 of 4 similarly priced models
OTIS Mock AIME 2024-2025
Best score per model · Higher is better
#23 of 158
- 98.9%Gemini 3.8 FlashNearby rank
- 98.9%GPT-6 LunaNearby rank
- 98.6%DeepSeek V4 Pro 0813
- 98.3%Claude Opus 4.8Nearby rank
- 98.3%DeepSeek V4.1 FlashNearby rank
- 98.3%Mistral Large 4Similar price
- 95.6%Gemini 3 Flash PreviewSimilar price
- 95.6%Kimi K2.7 CodeSimilar price
- 86.7%Qwen3 235B A22B Thinking 2507Similar price
Ahead of 4 of 4 similarly priced models
SimpleQA Verified
Best score per model · Higher is better
#25 of 87
- 66.8%Gemini 3 Flash PreviewSimilar price
- 55.8%Qwen3.7 MaxNearby rank
- 53.0%Claude Opus 4.8Nearby rank
- 52.9%DeepSeek V4 Pro 0813
- 52.0%
Qwen3.6 Max PreviewNearby rank - 51.7%Claude Opus 4.7Nearby rank
- 40.4%Qwen3 235B A22B Thinking 2507Similar price
- 36.5%Kimi K2.7 CodeSimilar price
- 20.0%Mistral Large 4Similar price
Ahead of 3 of 4 similarly priced models
Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.
Reasoning effort
ARC-AGI-2
Score by reasoning effort · Higher is better
All published scores
- Arena TextArena #591464
- high146495% CI 1457 to 1471 · 9,584 votes
- Arena WebDevArena #441581
- high158195% CI 1572 to 1591 · 4,604 votes
- ARC-AGI-2#33 of 8561.3%
- none0.8%
- low56.3%
- high59.7%
- max61.3%
- FrontierMath-Tier-4-v2-Private#38 of 7226.8%
- max26.8%
Source:Epoch AI evaluation - FrontierMath-Tiers-1-3-v2-Private#38 of 8764.6%
- max64.6%
Source:Epoch AI evaluation - GPQA diamond#28 of 16191.7%
- max91.7%
Source:Epoch AI evaluation - OTIS Mock AIME 2024-2025#23 of 15898.6%
- max98.6%
Source:Epoch AI evaluation - SimpleQA Verified#25 of 8752.9%
- max52.9%
Source:Epoch AI evaluation
Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.
Arena ratings: Arena (lmarena-ai/leaderboard-dataset), licensed CC BY 4.0. Ratings come from pairwise human votes and are only comparable within one leaderboard; the interval is Arena's 95% confidence interval. Model names are matched to the Vertical Tabs catalog.